利用采样数据间的相互关系,以电网元件相关先验概率为基础,通过D-S(Dempster-Shafer)证据理论对检测的故障信息进行融合,获取缺失数据发生的概率;同时,将概率的概念引入Petri网建模,将获取的状态概率值代入概率Petri网故障诊断模型进行故障的诊断,以解决电网故障诊断过程中的信息不完备问题.实例证明,该方法可扩充电网故障的诊断范围,保障电力系统安全运行.
The method of information fusion is proposed. The fusion process is treated by D-S (Dempster-Shafer) evidence decision rule. The occurred probability of the missed datum is obtained by the fault occurred prior-probability of the equipment and the relationship of the sample data. Then the ob- tained probability value can be applied to the probability Petri nets fault diagnostic model to diagnose the fault. The method can eliminate the influence of incomplete sample data in grid fault diagnosis. The proposed technique is proved to be availability by an example, and it can extend the application of fault diagnostic method for the electrical operation safety.